Software defined radio aims to bury GSM-CDMA hatchet

Mobile phone technology was supposed to give users the freedom of connectivity no matter where they happened to be, but the promise of global roaming – of being able to make wireless calls anywhere in the world – has not come to fruition.

At least, not yet. But a new technology that translates different kinds of transmissions could turn the global roaming dream into reality.

For the moment, untethered connectivity is anything but global, in part because mobile service providers can’t seem to agree on one type of technology. Some carriers employ code division multiple access (CDMA) networks to send calls to and from their destinations. For the most part others opt for global system for mobile communication (GSM) networks.

This patchwork of protocols makes for less-than-seamless wireless coverage. CDMA handset owners cannot make calls where GSM is the protocol of choice – areas such as the United Kingdom. If you happen to be a Bell Mobility or Telus Mobility customer here in Canada, you can kiss goodbye the notion of visiting England and calling someone on your mobile phone. Bell and Telus use CDMA technology, so their customers can’t connect in the U.K.

That is, they can’t today. The future, however, spells hope for global roaming, thanks to technology known as software defined radio (SDR). SDR essentially teaches networks to understand wireless languages other than their own, so GSM networks could send and receive CDMA transmissions and vice versa.

SDR means cellular customers could make calls anywhere wireless access exists, no matter what kind of handsets they have, and never mind the technology in the network.

“Global roaming is one of the benefits of SDR,” said Manuel Uhm, senior manager of strategic marketing with Burnaby, B.C.’s Spectrum Signal Processing Inc., a company focused on bringing SDR to the commercial wireless space. “Theoretically, you could have base stations that support CDMA phones as well as GSM phones.”

As well, SDR could spell a shift in the way enterprise telecom directors tackle mobile phone purchases – a movement away from technology concerns and towards more tangible factors.

Today when corporations go shopping for wireless handsets, they consider what sort of network the phones are designed to work with as well as network coverage. But with SDR, neither the handset nor the infrastructure matters. Enterprises could focus on the important stuff like prices and features.

“As a telecom manager of the future, I can afford to pay less attention to the underlying infrastructure,” said Warren Chaisatien, analyst with IDC Canada Ltd. in Toronto.

Still, it’s not as if SDR will affect corporate buying decisions in the near term. The technology has a long road ahead before hitting the big time.

Uhm said carriers interested in deploying SDR base stations would have to swap out entire networks in favour of software-defined infrastructure – an expensive and time-consuming undertaking.

However, he also pointed out that SDR could lead to cost savings as carriers migrate from second-generation (2G) networks to the vaunted third-generation (3G) infrastructure, which promises faster wireless data transmissions.

“[Carriers have] 2G base stations and now all of a sudden they have to go to 3G base stations,” Uhm said. “In order to do that, they have to put in an entirely new network. That’s going to cost billions of dollars…

“With software defined radio base stations, you have the ability to upgrade the protocol stack as you see fit, to support a mix of 2G and 3G, to support entirely 3G or entirely 2G or 3G and to change that mix as your subscriber mix changes.”

David Neale, vice-president of new product development with GSM mobile access provider Rogers AT&T Wireless in Toronto, said SDR has some merits, in theory. But in practice the technology is complicated, he said.

“At the 150,000-foot level, it all makes a huge amount of sense. But as you get closer to the ground, you realize those things below are actually mountains.”

Neale wondered just which wireless service providers would benefit from SDR. The company he works for might not be one of them.

Rogers’ GSM phones – handsets from Nokia Corp., Motorola Inc. and others – operate in three frequencies to account for various versions of GSM around the world. Neale said the carrier might add a tetraband handset to its offerings to further increase coverage.

He pointed out that GSM is more prevalent than CMDA. Although it doesn’t blanket the entire world, GSM comes awfully close to being a de facto global standard in Neale’s mind.

So why should Rogers bother with SDR?

“If I have tetraband GSM…that pretty much addresses the world for me. My desire to roam onto a CDMA network is pretty thin. One might argue, in fact, that the people who would best benefit are the minor standards,” carriers using CDMA for example.

Brian O’Shaughnessy, vice-president, wireless technology with Bell Mobility (a CDMA operator), said improved integration between various mobile standards would indeed benefit CDMA customers. However, he’s not sure SDR is the way to get there.

O’Shaughnessy said carriers could use software to teach base stations other languages. But it would be “so complex” an operation that “you’d rather build two networks.”

As far as he’s concerned, SDR belongs in the handset, the device that’s doing the roaming, not the base station. And even that isn’t the only option, O’Shaughnessy said, pointing out that in the future manufacturers will build handsets capable of transmitting GSM and CDMA thanks to fancy hardware, not necessarily software.

Chaisatien from IDC Canada said the network-agnostic future might come to pass without the aid of SDR.

“Towards the end of the decade, there’s a lot of speculation that these two standards (CMDA and GSM) will converge,” he said. “But nobody is certain about that.”

If convergence is in the cards for wireless networks, “it would render the software [defined radio] meaningless,” Chaisatien said. However, he added that it’s too soon to predict how any CDMA-GSM convergence would shake out. Meanwhile, the analyst said CDMA and GSM operators would remain at odds well past 2005 and likewise well past the advent of 3G.

Other hurdles could hinder SDR’s deployment. Consider the standard itself, which, judging by the experiences of one industry insider, is akin to a battleground.

Steve Bernier, project leader, military satellite communications with the Communications Research Centre (CRC) in Ottawa, said some members of the Software Defined Radio Forum (SDRF) – a non-profit organization bent on bringing SDR to wireless – are fighting for control of the protocol.

Bernier said the U.S. military, which is interested in radio signal interoperability (see sidebar), stands behind a version of SDR called software communication architecture (SCA v2.2). This specification is supposed to be the basic starting point for SDR designs, but Bernier said some SDRF members don’t see it that way.

“We’ve been participating in the forum for three years, and all we ever see is the big guys trying to impose their own standards or own interpretations of the standard.”

The CRC stepped up to stop the infighting with an SDR reference plan, a framework that the 100 or so forum members can work with and derive recommendations from.

“We think the way to do it is adopt the standard and if you don’t like it, submit a change proposal,” Bernier said. “It will be reviewed and adopted internationally. You’ll get acceptance from all of the partners, instead of trying to impose your own standard.”

Bernier said SDR would come online in six or seven years, but Uhm from Spectrum was more optimistic.

“It’s getting more momentum now. I’d say it’s only four years away before you’ll see the majority of base stations deployed being software defined.”